brAInfood: How ‘trustworthy’ is artificial intelligence?

Date of publication: March 2020 / Checked if updates were needed: March 2024

WHAT IS THIS BRAINFOOD ABOUT?

This brAInfood summarizes the 7 requirements of artificial intelligence (AI), written by the European Commission’s High-Level Expert Group on AI (AI HLEG). For each requirement, a number of questions are formulated that enable you to reflect on the trustworthiness of your AI system. This should allow you to have a better understanding of what ‘trustworthy’ AI means and how you can use your AI system ‘in a responsible way’.

WHY SHOULD I READ THIS?

If you program, implement, manage or work with an AI system, this brAInfood can help you to explore how well-adapted the system is to the European Commission’s standards on ethical AI. Also, if you interact with an AI system in your daily life (e.g. smartphone assistants, recommendation systems, …), this brAInfood can help you to critically think about the implications of this system.

With this brAInfood, the Knowledge Centre Data & Society hopes to foster the debate on artificial intelligence and aims to provide guidance on the development and implementation of ‘trustworthy’ AI. This brAInfood raises important questions about the challenges in operationalizing ethical values and societal norms when developing and implementing AI systems. By keeping in mind these questions during the development and implementation phase, you are able to continuously reflect on the trustworthiness of your AI system.

If you require a more detailed approach, the Knowledge Centre Data & Society recently developed the AI Blindspots cards. By using this tool, you proactively reflect on the (ethical and social) decisions and actions you want to take at the start of your project or during the development of your AI system.

WHAT CAN I FIND OUT WHEN READING THIS?

This brAInfood from the Data & Society Knowledge Centre includes:

• A concise but clear set of questions for each of the 7 requirements of artificial intelligence (AI), written by the European Commission’s High-Level Expert Group on AI (AI HLEG), to test if your AI system is in accordance with these guidelines.